DIAB-6247; No. of Pages 5 diabetes research and clinical practice xxx (2015) xxx.e1–xxx.e5
Contents available at ScienceDirect
Diabetes Research and Clinical Practice journ al h ome pa ge : www .elsevier.co m/lo cate/diabres
Lower serum 25 (OH) D concentrations in type 1 diabetes: A meta-analysis Rennan Feng a,b, Yanchuan Li a,b, Guoqiang Li c, Zhen Li a,b, Yingfeng Zhang a,b, Qiang Li d,*, Changhao Sun a,b,* a
Department of Nutrition and Food Hygiene, School of Public Health, Harbin Medical University, Harbin, China National Key Discipline, Harbin Medical University, Harbin, China c Dean Office, Harbin Medical University, Harbin, China d Department of Endocrinology, Second Affiliated Hospital of Harbin Medical University, Harbin, China b
article info
abstracts
Article history:
Objective: Vitamin D may have protective effects against type 1 diabetes (T1D). However, the
Received 23 April 2014
results of current cross-sectional studies have been inconsistent. We aimed to examine the
Received in revised form
association of serum vitamin D and T1D by conducting a meta-analysis.
28 October 2014
Methods: Multiple databases were searched until April 2014 to identify relevant studies.
Accepted 29 December 2014
The search term was ‘‘serum vitamin D’’ and ‘‘type 1 diabetes’’. Random effects model was
Available online xxx
used to calculate the pooled risk estimate. The data were accessed by software Review manager 5.2.
Keywords:
Results: We included a total of 23 studies (11 studies with 1900 participants of children or
Type 1 diabetes
adolescents and 13 studies with 3494 participants of adults) in our meta-analysis. In children
Serum vitamin D
or adolescents, serum 25(OH) D in T1D was 5.69 ng/ml [95% confidence interval (CI) (2.82–
Meta-analysis
8.55, P < 0.0001) lower than in healthy controls. In adults, serum 25(OH) D in T1D was 2.61 ng/ml [95% confidence interval (CI) (1.13–4.09, P = 0.0005) lower than in healthy controls. We did not find evidence of publication bias. Conclusion: Serum 25OHD is significantly lower in patients with T1D than in healthy controls. # 2015 Elsevier Ireland Ltd. All rights reserved.
1.
Introduction
Type 1 diabetes (T1D) is one of the most common lifethreatening endocrine disorders in children and young adults worldwide and its incidence appears to be increasing during the past decades [1,2]. It is characterized by autoimmune destruction of insulin-producing b cells in the pancreas but the mechanisms of this autoimmune process are still not well known.
Vitamin D can be produced endogenously by skin when exposed to sunlight and also can be obtained from foods or supplements. In northern countries, little vitamin D is produced from the skin during winter months, because of short time spent outside, light skin color, longer winter and thick clothes [3]. Ecological studies had shown an association between low ultraviolet B irradiance and high incidence rates of T1D in children, leading to the hypothesis that vitamin D may play a role in the process of T1D [4].
* Corresponding authors at: Harbin Medical University, Department of Nutrition and Food Hygiene, School of Public Health, Harbin 150081, Heilongjiang, China. Tel.: +86 0451 87502801; fax: +86 0451 87502885. E-mail addresses: [email protected] (Q. Li), [email protected] (C. Sun). http://dx.doi.org/10.1016/j.diabres.2014.12.008 0168-8227/# 2015 Elsevier Ireland Ltd. All rights reserved.
Please cite this article in press as: Feng R, et al. Lower serum 25 (OH) D concentrations in type 1 diabetes: A meta-analysis. Diabetes Res Clin Pract (2015), http://dx.doi.org/10.1016/j.diabres.2014.12.008
DIAB-6247; No. of Pages 5
e2
diabetes research and clinical practice xxx (2015) xxx.e1–xxx.e5
Vitamin D deficiency/insufficiency is a public health issue, especially in developing country [5]. In China, a study in Beijing showed more than 40% of school-aged girls had 25(OH) D levels below 12.5 nmol/L [6]. More seriously, a large percentage of had pregnant women in Nanjing had serum 25(OH) D levels below 25 nmol/L [7] and also in pregnant women with gestational diabetes mellitus in Beijing [8]. Vitamin D deficiency/insufficiency in Chinese represents a significant but modifiable public health risk that deserves greater attention. Since a report of a reduced risk of developing T1D by vitamin D supplementation during the first year of life [9], several studies reported a high prevalence of vitamin D deficiency in patients with T1D [10–12]. However, subsequent studies reported inconsistent results [13]. As small sample size may have led to a weak or non-significant associations in those studies, the aim of this meta-analysis was to examine the association between serum vitamin D and risk of T1D in children or adolescents and adults.
2.
Methods
We generally followed the Meta-analysis of Observational Studies in Epidemiology (MOOSE) guidelines [14]. We searched literature published in PubMed, EBSCO, and Springer database until April 2014 to identify relevant studies that assessed the association between serum vitamin D and T1D. Search terms included ‘‘vitamin D’’ and ‘‘type 1 diabetes’’.
Because there were no cohort studies, we could only select and analyze case-control study which reported serum vitamin D concentrations of T1DM patients and age-sex-matched controls. Two authors (R-N. F. and Y-C. L.) extracted data independently and disagreements were resolved by discussion. Information on each study is provided in Table 1. In the meta-analysis graphical representation, the area of the black square indicates the weight contributed by each study. Heterogeneity was estimated by I2, substantial heterogeneity exists when I2 exceeds 50%. A random effect model was used if heterogeneity was observed, while the fixed-effect model was applied in the absence of heterogeneity. All analyses were performed using Review Manager 5.2 software (Cochrane Collaboration, Oxford, England). A P value of less than 0.05 was considered to be statistically significant.
3.
Results
Fig. 1 presents a flow chart of the study selection process. We initially identified a total of 113 articles. After screening titles and abstracts, reviews and irrelevant articles, 40 relevant articles remained for full-text evaluation. 17 articles were excluded for the following reasons: no control group (n = 9), no 25OHD concentrations reported (n = 6), cohort or intervention studies (n = 2). Finally 11 studies of children and adolescents [10,15–23] and 13 studies of adults [24–35] were included in our study. Characteristics of the selected studies are presented in Table 1.
Table 1 – Characteristics of included studies. Author
Year
Country
Control Number Male Age
Children or adolescents 1985 Norway Rodland Verrotti 1988 Italy 2008 UK Bener 2011 Saudi Arabia Bin-Abbas 2011 Egypt Hamed Daga 2012 India 2012 Australia Greer
191 21 170 100 15 41 46
Oilinki Franchi Somia
2012 2013 2014
Finland Italy Egypt
310 166 120
Adults Hampson Verrotti Fassbender Bierschenk
1998 1999 2002 2009
UK Italy Germany America
20 24 57 153
Singh Singh Masse Yavuz Devaraj Thrailkill Gorham Al-Daghri
2009 2009 2010 2010 2011 2011 2012 2014
UK UK Canada Turkey America America America Saudi Arabia
25 41 32 134 36 55 1000 60
Case 25(OH)D(ng/ml) Number Male Age
88
10.5
7 25
8.47 16.98 9.1
32.0 15.5 22.09 5.01 18.45 9.56 17.9 5.6 34.2 2.3 16.64 7.83 36.6 8.2
92 48
8.7 11.1
20.6 8.2 19.5 98.7 18.6 5.4
0
41.8
27 69
33.5 5–65.1
12 20 0 61 14 31
43.8 43.1 34.8 26.2 31 22.3
32
36.7
24.5 13 38.1 4.9 22.21 15.19 20.1 32.9 27.1 27.7 27.6 29.2 31.4 12.1 35.1 24 32.0 9.0 34.0 85.1 29.0 13.2 13.4 0.6
85 21 170 100 36 13 42 14 15 58 120
31 24 75 46 110 35 40 27 117 24 115 1000 60
80
9.9
13 6 21 7 5 32 42
10.38 15.15 12.9 7.5 7.3 9.2 11.7
0
42.4
42 23 60 20 20 0 57 8 61
33.5 5.9–35.0 5.1–62.6 44.2 42.4 36.9 27.6 33 17.5
34
25.9
P 25(OH)D(ng/ml) 23.1 9.2 24.33 6.04 15.80 9.23 14.7 5.7 18.7 5.1 11.36 4.74 32.6 7.7 28.2 20.8 12.7 2.6 14.5 61.3 14.2 5.0
Contents available at ScienceDirect
Diabetes Research and Clinical Practice journ al h ome pa ge : www .elsevier.co m/lo cate/diabres
Lower serum 25 (OH) D concentrations in type 1 diabetes: A meta-analysis Rennan Feng a,b, Yanchuan Li a,b, Guoqiang Li c, Zhen Li a,b, Yingfeng Zhang a,b, Qiang Li d,*, Changhao Sun a,b,* a
Department of Nutrition and Food Hygiene, School of Public Health, Harbin Medical University, Harbin, China National Key Discipline, Harbin Medical University, Harbin, China c Dean Office, Harbin Medical University, Harbin, China d Department of Endocrinology, Second Affiliated Hospital of Harbin Medical University, Harbin, China b
article info
abstracts
Article history:
Objective: Vitamin D may have protective effects against type 1 diabetes (T1D). However, the
Received 23 April 2014
results of current cross-sectional studies have been inconsistent. We aimed to examine the
Received in revised form
association of serum vitamin D and T1D by conducting a meta-analysis.
28 October 2014
Methods: Multiple databases were searched until April 2014 to identify relevant studies.
Accepted 29 December 2014
The search term was ‘‘serum vitamin D’’ and ‘‘type 1 diabetes’’. Random effects model was
Available online xxx
used to calculate the pooled risk estimate. The data were accessed by software Review manager 5.2.
Keywords:
Results: We included a total of 23 studies (11 studies with 1900 participants of children or
Type 1 diabetes
adolescents and 13 studies with 3494 participants of adults) in our meta-analysis. In children
Serum vitamin D
or adolescents, serum 25(OH) D in T1D was 5.69 ng/ml [95% confidence interval (CI) (2.82–
Meta-analysis
8.55, P < 0.0001) lower than in healthy controls. In adults, serum 25(OH) D in T1D was 2.61 ng/ml [95% confidence interval (CI) (1.13–4.09, P = 0.0005) lower than in healthy controls. We did not find evidence of publication bias. Conclusion: Serum 25OHD is significantly lower in patients with T1D than in healthy controls. # 2015 Elsevier Ireland Ltd. All rights reserved.
1.
Introduction
Type 1 diabetes (T1D) is one of the most common lifethreatening endocrine disorders in children and young adults worldwide and its incidence appears to be increasing during the past decades [1,2]. It is characterized by autoimmune destruction of insulin-producing b cells in the pancreas but the mechanisms of this autoimmune process are still not well known.
Vitamin D can be produced endogenously by skin when exposed to sunlight and also can be obtained from foods or supplements. In northern countries, little vitamin D is produced from the skin during winter months, because of short time spent outside, light skin color, longer winter and thick clothes [3]. Ecological studies had shown an association between low ultraviolet B irradiance and high incidence rates of T1D in children, leading to the hypothesis that vitamin D may play a role in the process of T1D [4].
* Corresponding authors at: Harbin Medical University, Department of Nutrition and Food Hygiene, School of Public Health, Harbin 150081, Heilongjiang, China. Tel.: +86 0451 87502801; fax: +86 0451 87502885. E-mail addresses: [email protected] (Q. Li), [email protected] (C. Sun). http://dx.doi.org/10.1016/j.diabres.2014.12.008 0168-8227/# 2015 Elsevier Ireland Ltd. All rights reserved.
Please cite this article in press as: Feng R, et al. Lower serum 25 (OH) D concentrations in type 1 diabetes: A meta-analysis. Diabetes Res Clin Pract (2015), http://dx.doi.org/10.1016/j.diabres.2014.12.008
DIAB-6247; No. of Pages 5
e2
diabetes research and clinical practice xxx (2015) xxx.e1–xxx.e5
Vitamin D deficiency/insufficiency is a public health issue, especially in developing country [5]. In China, a study in Beijing showed more than 40% of school-aged girls had 25(OH) D levels below 12.5 nmol/L [6]. More seriously, a large percentage of had pregnant women in Nanjing had serum 25(OH) D levels below 25 nmol/L [7] and also in pregnant women with gestational diabetes mellitus in Beijing [8]. Vitamin D deficiency/insufficiency in Chinese represents a significant but modifiable public health risk that deserves greater attention. Since a report of a reduced risk of developing T1D by vitamin D supplementation during the first year of life [9], several studies reported a high prevalence of vitamin D deficiency in patients with T1D [10–12]. However, subsequent studies reported inconsistent results [13]. As small sample size may have led to a weak or non-significant associations in those studies, the aim of this meta-analysis was to examine the association between serum vitamin D and risk of T1D in children or adolescents and adults.
2.
Methods
We generally followed the Meta-analysis of Observational Studies in Epidemiology (MOOSE) guidelines [14]. We searched literature published in PubMed, EBSCO, and Springer database until April 2014 to identify relevant studies that assessed the association between serum vitamin D and T1D. Search terms included ‘‘vitamin D’’ and ‘‘type 1 diabetes’’.
Because there were no cohort studies, we could only select and analyze case-control study which reported serum vitamin D concentrations of T1DM patients and age-sex-matched controls. Two authors (R-N. F. and Y-C. L.) extracted data independently and disagreements were resolved by discussion. Information on each study is provided in Table 1. In the meta-analysis graphical representation, the area of the black square indicates the weight contributed by each study. Heterogeneity was estimated by I2, substantial heterogeneity exists when I2 exceeds 50%. A random effect model was used if heterogeneity was observed, while the fixed-effect model was applied in the absence of heterogeneity. All analyses were performed using Review Manager 5.2 software (Cochrane Collaboration, Oxford, England). A P value of less than 0.05 was considered to be statistically significant.
3.
Results
Fig. 1 presents a flow chart of the study selection process. We initially identified a total of 113 articles. After screening titles and abstracts, reviews and irrelevant articles, 40 relevant articles remained for full-text evaluation. 17 articles were excluded for the following reasons: no control group (n = 9), no 25OHD concentrations reported (n = 6), cohort or intervention studies (n = 2). Finally 11 studies of children and adolescents [10,15–23] and 13 studies of adults [24–35] were included in our study. Characteristics of the selected studies are presented in Table 1.
Table 1 – Characteristics of included studies. Author
Year
Country
Control Number Male Age
Children or adolescents 1985 Norway Rodland Verrotti 1988 Italy 2008 UK Bener 2011 Saudi Arabia Bin-Abbas 2011 Egypt Hamed Daga 2012 India 2012 Australia Greer
191 21 170 100 15 41 46
Oilinki Franchi Somia
2012 2013 2014
Finland Italy Egypt
310 166 120
Adults Hampson Verrotti Fassbender Bierschenk
1998 1999 2002 2009
UK Italy Germany America
20 24 57 153
Singh Singh Masse Yavuz Devaraj Thrailkill Gorham Al-Daghri
2009 2009 2010 2010 2011 2011 2012 2014
UK UK Canada Turkey America America America Saudi Arabia
25 41 32 134 36 55 1000 60
Case 25(OH)D(ng/ml) Number Male Age
88
10.5
7 25
8.47 16.98 9.1
32.0 15.5 22.09 5.01 18.45 9.56 17.9 5.6 34.2 2.3 16.64 7.83 36.6 8.2
92 48
8.7 11.1
20.6 8.2 19.5 98.7 18.6 5.4
0
41.8
27 69
33.5 5–65.1
12 20 0 61 14 31
43.8 43.1 34.8 26.2 31 22.3
32
36.7
24.5 13 38.1 4.9 22.21 15.19 20.1 32.9 27.1 27.7 27.6 29.2 31.4 12.1 35.1 24 32.0 9.0 34.0 85.1 29.0 13.2 13.4 0.6
85 21 170 100 36 13 42 14 15 58 120
31 24 75 46 110 35 40 27 117 24 115 1000 60
80
9.9
13 6 21 7 5 32 42
10.38 15.15 12.9 7.5 7.3 9.2 11.7
0
42.4
42 23 60 20 20 0 57 8 61
33.5 5.9–35.0 5.1–62.6 44.2 42.4 36.9 27.6 33 17.5
34
25.9
P 25(OH)D(ng/ml) 23.1 9.2 24.33 6.04 15.80 9.23 14.7 5.7 18.7 5.1 11.36 4.74 32.6 7.7 28.2 20.8 12.7 2.6 14.5 61.3 14.2 5.0
